convnext-base-3e-4
This model is a fine-tuned version of facebook/convnextv2-base-22k-384 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.3332
- Accuracy: 0.9312
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0003
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.8947 | 1.0 | 1099 | 0.7364 | 0.7885 |
| 0.7643 | 2.0 | 2198 | 0.6286 | 0.8171 |
| 0.6036 | 3.0 | 3297 | 0.5258 | 0.8481 |
| 0.5012 | 4.0 | 4396 | 0.4911 | 0.8696 |
| 0.3926 | 5.0 | 5495 | 0.3804 | 0.8930 |
| 0.3348 | 6.0 | 6594 | 0.4132 | 0.8970 |
| 0.2594 | 7.0 | 7693 | 0.3627 | 0.9153 |
| 0.1751 | 8.0 | 8792 | 0.3507 | 0.9308 |
| 0.1613 | 9.0 | 9891 | 0.3488 | 0.9300 |
| 0.1102 | 10.0 | 10990 | 0.3485 | 0.9280 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.1.2
- Datasets 2.18.0
- Tokenizers 0.15.2